Package org.vietspider.token.attribute

Examples of org.vietspider.token.attribute.Attributes

43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
      if(node.isNode(Name.FORM)) {
//        System.out.println("== > thay "+ node + " : "+ node.isOpen()  );
        if(node.getType() == TypeToken.TAG) {
          if(form != null) forms.add(form);
          form = new Form();
          Attributes attributes = node.getAttributes();
          Attribute attribute = attributes.get(NAME_ATTR);
          if(attribute != null) form.setName(attribute.getValue());
          attribute = attributes.get("method");
          if(attribute != null) form.setMethod(attribute.getValue());
          attribute = attributes.get(ACTION);
          if(attribute != null) form.setAction(attribute.getValue());
        } else {
          if(form != null) forms.add(form);
          form = null;
        }
      } else if(node.isNode(Name.INPUT)
          || node.isNode(Name.TEXTAREA)
          || node.isNode(Name.SELECT)) {       
        if(form == null) continue;
       
        Attributes attributes = node.getAttributes();
        Attribute attribute = attributes.get(NAME_ATTR);
        if(attribute == null) attribute = attributes.get(ID_ATTR);;
        if(attribute == null) continue;
        String name  = attribute.getValue();
        if(name == null || name.trim().isEmpty()) continue;
       
        attribute = attributes.get(TYPE_ATTR);
        String type = null;
        if(attribute != null) type = attribute.getValue();
        if(type == null) type = "text";
        type = type.trim();
       
        if(type.equalsIgnoreCase("submit")) {
          if(name.toLowerCase().indexOf("cancel") > -1
              || name.toLowerCase().indexOf("reset") > -1
              || name.toLowerCase().indexOf("clear") > -1) continue;
        }
       
        if("radio".equalsIgnoreCase(type)
            || "checkbox".equalsIgnoreCase(type)) {
          attribute = attributes.get("checked");
          if(attribute == null) continue;
        }
       
        attribute = attributes.get(VALUE);
        String value = null;
        if(attribute != null) value = attribute.getValue();
        if(value == null) value = "";
        value  = value.trim();
       
View Full Code Here
42
43
44
45
46
47
48
49
50
51
52
53
    NodeIterator iterator = root.iterator();
  
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(!node.isNode(Name.DIV)) continue;
      Attributes attributes = node.getAttributes();
      Attribute attribute = attributes.get("id");
      if(attribute == null) continue;
      String value = attribute.getValue();
      if(value == null) continue;
     
      if("posts".equalsIgnoreCase(value)) {
View Full Code Here
147
148
149
150
151
152
153
154
155
156
157
158
  private HTMLNode searchUserNode(HTMLNode root) {
    NodeIterator iterator = root.iterator();
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(!node.isNode(Name.A)) continue;
      Attributes attributes = node.getAttributes();
      Attribute attribute = attributes.get("class");
      if(attribute == null) continue;
      String value = attribute.getValue();
      if(value == null) continue;
      value = value.toLowerCase();
      if(value.indexOf("bigusername") > -1) return node;
View Full Code Here
163
164
165
166
167
168
169
170
171
172
173
174
  private HTMLNode searchContentNode(HTMLNode root, String clazz) {
    NodeIterator iterator = root.iterator();
    while(iterator.hasNext()) {
      HTMLNode node = iterator.next();
      if(!node.isNode(Name.DIV)) continue;
      Attributes attributes = node.getAttributes();
      Attribute attribute = attributes.get("id");
      if(attribute == null) continue;
      String value = attribute.getValue();
      if(value == null) continue;
      value = value.toLowerCase();
      if(value.indexOf(clazz) > -1) {//"post_message"
View Full Code Here
181
182
183
184
185
186
187
188
189
190
191
192
  private boolean isPageList(HTMLNode node) {
    NodeIterator iterator = node.iterator();
    while(iterator.hasNext()) {
      HTMLNode n = iterator.next();
      if(!n.isNode(Name.A)) continue;
      Attributes attributes = n.getAttributes();
      Attribute attribute = attributes.get("href");
      if(attribute == null) continue;
      String value = attribute.getValue();
      if(value == null) continue;
      if(value.indexOf("page=") > -1) return true;
    }
View Full Code Here
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
    for(;i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
     
      if(node.isNode(Name.FORM)) {
        if(formName == null) break;       
        Attributes attrs = node.getAttributes();
        String name = getAttribute(attrs, NAME_ATTR);
        if(formName.equalsIgnoreCase(name)) break;
      }
    }

    HTMLNode form = null;
    List<HTMLNode> inputs = new ArrayList<HTMLNode>();

    for(; i < tokens.size(); i++) {
      NodeImpl node = tokens.get(i);
      if(node.isNode(Name.FORM)) {
        if(node.isOpen()) {
          form = node;
        } else {
          break;
        }
      } else if(node.isNode(Name.INPUT)
          || node.isNode(Name.TEXTAREA)
          || node.isNode(Name.SELECT)) {       
        inputs.add(node);
      }
    }
   
    if(form == null || inputs.size() < 1) {
      message = "Error: form not found";
      return ERROR;
    }
   
    String address = getAttribute(form, ACTION);
    if(address == null) {
      message = "Error: post address not found!";
      return ERROR;
    }
   
    List<NameValuePair> params = new ArrayList<NameValuePair>();
    for(i = 0 ; i < inputs.size(); i++) {
      Attributes attrs = inputs.get(i).getAttributes();
      String name = getAttribute(attrs, NAME_ATTR);
      if(name == null) getAttribute(attrs, ID_ATTR);
      if(name == null) continue;
      String type = getAttribute(attrs, TYPE_ATTR);
      if(type == null) type = "text";
View Full Code Here
267
268
269
270
271
272
273
274
275
276
//    }
    params.add(newPair);
  }

  public String getAttribute(HTMLNode node, String name) {
    Attributes attrs = node.getAttributes();
    int idx = attrs.indexOf(name);
    if(idx < 0) return null;
    return attrs.get(idx).getValue();
  }
View Full Code Here
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
 
  private boolean equalsFormat(HTMLNode node1, HTMLNode node2) {
    Name name1 = node1.getName();
    Name name2 = node2.getName();
    if(name1 != name2) return false;
    Attributes attributes1 = node1.getAttributes();
    Attributes attributes2 = node2.getAttributes();
    if(attributes1.size() != attributes2.size()) return false;
    for(int i = 0;  i < attributes1.size(); i++) {
      Attribute attribute1 = attributes1.get(i);
      Attribute attribute2 = attributes2.get(i);
      if(attribute1 == null && attribute2 != null)  return false;
      if(attribute1 != null && attribute2 == null)  return false;
      if(attribute1 != null && attribute2 != null
          && !attribute1.getName().equalsIgnoreCase(attribute2.getName()))  return false;
    }
View Full Code Here
14
15
16
17
18
19
20
21
* Feb 9, 2009 
*/
public class RenderNodeUtils {
 
  public static int getIntAttrValue(HTMLNode node, String name) {
    Attributes attributes = node.getAttributes();
    return getIntAttrValue(attributes, name);
  }
View Full Code Here
85
86
87
88
89
90
91
92
93
94
95
        if(attrs == null || attrs.length < 1) {
//        System.out.println(" da xay ra roi ");
          htmlValues.add(htmlChildren.get(i));
        } else {
//          System.out.println(" xay ra ");
          Attributes nodeAttributes = htmlChildren.get(i).getAttributes();
          if(matcher.contains(nodeAttributes, attrs)) htmlValues.add(htmlChildren.get(i));
        }
      }
      counter++;
    } 
View Full Code Here
TOP

Related Classes of org.vietspider.token.attribute.Attributes

  • AttributeExample
  • AttributeExplorer
  • JavascriptExtractor
  • org.vietspider.browser.form.FormUtils
  • org.vietspider.browser.LoginWebsiteHandler
  • org.vietspider.browser.VirtualFormPost
  • org.vietspider.browser.VirtualFormPostBak
  • org.vietspider.html.path2.HTMLExtractor
  • org.vietspider.html.path2.LookupNode
  • org.vietspider.html.renderer.checker.ContentChecker
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.